Feature propagation on image webs for enhanced image retrieval. Brachmann, E., Spehr, M., & Gumhold, S. In abstract bibtex The bag-of-features model is often deployed in content-based image retrieval to measure image similarity. In cases where the visual appearance of semantically similar images differs largely, feature histograms mismatch and the model fails. We increase the robustness of feature histograms by automatically augmenting them with features of related images. We establish image relations by image web construction and adapt a label propagation scheme from the domain of semi-supervised learning for feature augmentation. While the benefit of feature augmentation has been shown before, our approach refrains from the use of semantic labels. Instead we show how to increase the performance of the bag-of-features model substantially on a completely unlabeled image corpus.
@inproceedings{ Brachmann-2013-FPI,
author = {Eric Brachmann and
Marcel Spehr and
Stefan Gumhold},
title = {Feature propagation on image webs for enhanced image retrieval},
abstract = {
The bag-of-features model is often deployed in content-based image retrieval to measure image similarity.
In cases where the visual appearance of semantically similar images differs largely, feature histograms
mismatch and the model fails. We increase the robustness of feature histograms by automatically augmenting
them with features of related images. We establish image relations by image web construction and adapt a
label propagation scheme from the domain of semi-supervised learning for feature augmentation. While the
benefit of feature augmentation has been shown before, our approach refrains from the use of semantic
labels. Instead we show how to increase the performance of the bag-of-features model substantially on a
completely unlabeled image corpus.
}
}
Downloads: 0
{"_id":{"_str":"537a130fb3aa042605004640"},"__v":6,"authorIDs":[],"author_short":["Brachmann, E.","Spehr, M.","Gumhold, S."],"bibbaseid":"brachmann-spehr-gumhold-featurepropagationonimagewebsforenhancedimageretrieval","bibdata":{"downloads":0,"bibbaseid":"brachmann-spehr-gumhold-featurepropagationonimagewebsforenhancedimageretrieval","urls":{},"role":"author","abstract":"The bag-of-features model is often deployed in content-based image retrieval to measure image similarity. In cases where the visual appearance of semantically similar images differs largely, feature histograms mismatch and the model fails. We increase the robustness of feature histograms by automatically augmenting them with features of related images. We establish image relations by image web construction and adapt a label propagation scheme from the domain of semi-supervised learning for feature augmentation. While the benefit of feature augmentation has been shown before, our approach refrains from the use of semantic labels. Instead we show how to increase the performance of the bag-of-features model substantially on a completely unlabeled image corpus.","author":["Brachmann, Eric","Spehr, Marcel","Gumhold, Stefan"],"author_short":["Brachmann, E.","Spehr, M.","Gumhold, S."],"bibtex":"@inproceedings{ Brachmann-2013-FPI,\n author = {Eric Brachmann and\n Marcel Spehr and\n Stefan Gumhold},\n title = {Feature propagation on image webs for enhanced image retrieval},\n abstract = {\nThe bag-of-features model is often deployed in content-based image retrieval to measure image similarity. \nIn cases where the visual appearance of semantically similar images differs largely, feature histograms \nmismatch and the model fails. We increase the robustness of feature histograms by automatically augmenting\nthem with features of related images. We establish image relations by image web construction and adapt a \nlabel propagation scheme from the domain of semi-supervised learning for feature augmentation. While the \nbenefit of feature augmentation has been shown before, our approach refrains from the use of semantic \nlabels. Instead we show how to increase the performance of the bag-of-features model substantially on a\ncompletely unlabeled image corpus.\n}\n}","bibtype":"inproceedings","id":"Brachmann-2013-FPI","key":"Brachmann-2013-FPI","title":"Feature propagation on image webs for enhanced image retrieval","type":"inproceedings"},"bibtype":"inproceedings","biburl":"http://tu-dresden.de/die_tu_dresden/fakultaeten/fakultaet_informatik/smt/cgv/mitarbeiter/marcel_spehr/mspehr.bib","downloads":0,"keywords":[],"search_terms":["feature","propagation","image","webs","enhanced","image","retrieval","brachmann","spehr","gumhold"],"title":"Feature propagation on image webs for enhanced image retrieval","year":null,"dataSources":["E69SmTmC27rWx8zht"]}